Case Study: Surviving the Lunar Night — Designing a Surface Power System

"On the Moon, the hardest engineering problem is not the day. It is the two weeks of dark that follow."

Executive Summary

In the first case study we audited a finished spacecraft. Here we design one from a blank page, and we let the environment force the answer. Our task: power a small science station on the Moon's equator, where every lunar day is followed by a night roughly 354 hours long — about two Earth weeks of continuous darkness and brutal cold. We will try the "obvious" solution first — solar panels and a battery, the recipe for every Earth satellite — and watch it collapse under the mass of a battery big enough to store two weeks of energy. Then we will reach for the radioisotope generator of §25.4 and find it fits as if the problem were made for it. Finally we scale the station up to a crewed base and watch the same logic demand a fission reactor. This is the calculation that decides how humanity will live on other worlds, and it is nothing more than the array, battery, and RTG sizing of this chapter, pushed to an environment that punishes the wrong choice with certain death of the spacecraft.

Skills applied

• Sizing a battery for an extreme eclipse — the two-week lunar night (§25.3).
• Recognizing when battery mass makes solar-plus-storage infeasible (§25.3, theme 4).
• Selecting an RTG for a low-power, long-dark mission (§25.4).
• Escalating to fission for a high-power surface base (§25.5), and comparing sources by specific power.

Background

The requirement

• Load: $P = 100\ \text{W}$ continuous — a modest suite of instruments, a computer, a low-rate radio, and survival heaters, running day and night.
• Environment: the Moon's equator. A lunar day and night each last about 14.77 Earth days; the night is therefore

$$ T_{\text{night}} = 14.77 \times 24 \approx 354\ \text{hours}. $$

• Cold: in the long night the surface falls below $-170\,^\circ\text{C}$, so the survival heaters are not optional — losing power means freezing solid, permanently.
• Life: 10 years, unattended. No one is coming to swap a battery.

The night, not the day, is the enemy. Let us make it quantitative.

Phase 1: The "obvious" design — solar panels and a battery

Every Earth satellite runs on an array plus a battery, so try it. During the 354-hour lunar day the array charges a battery; during the 354-hour night the battery alone carries the $100\ \text{W}$ load. Size the battery by §25.3. The energy it must deliver across the night is

$$ E_{\text{night}} = P \times T_{\text{night}} = 100\ \text{W} \times 354\ \text{h} = 35{,}400\ \text{W·h} = 35.4\ \text{kW·h}. $$

That is already a startling number — a 100 W load, but 35 kilowatt-hours to nurse it through one night, because the night is so unimaginably long. Now apply the depth-of-discharge and efficiency penalties. Even being generous — deep cycling at $\text{DoD} = 0.80$ (there are only ~130 night-cycles in 10 years, so deep discharge is acceptable) and discharge efficiency $\eta = 0.90$ — the installed capacity is

$$ C = \frac{E_{\text{night}}}{\text{DoD}\times\eta} = \frac{35{,}400}{0.80 \times 0.90} = \frac{35{,}400}{0.72} \approx 49{,}200\ \text{W·h}. $$

At the best space lithium-ion energy density, $150\ \text{W·h/kg}$, that battery masses

$$ m_{\text{batt}} = \frac{49{,}200}{150} \approx 328\ \text{kg}. $$

💡 Intuition: Read that number and feel the absurdity. To keep a $100\ \text{W}$ load — about two household light bulbs — alive through the lunar night, you must land a third of a tonne of battery. Even in the impossible ideal case (use 100% of the battery, zero losses) it is $35{,}400/150 = 236\ \text{kg}$. The battery outmasses the entire rest of the station many times over. This is theme 4 — mass is the enemy — screaming: the two-week night makes stored energy ruinously heavy.

Phase 2: The array is not the problem — the battery is

To be fair to the solar design, size its array too. During the 354-hour day the array must both run the $100\ \text{W}$ load and recharge the $49{,}200\ \text{W·h}$ the night will consume. The recharge power, averaged over the day, is $49{,}200/354 \approx 139\ \text{W}$, so the array must supply roughly

$$ P_{\text{array}} \approx 100 + 139 = 239\ \text{W}. $$

On the Moon's equator at midday the flux is the full $1361\ \text{W/m}^2$ (no atmosphere), and averaged over the slow day the sun angle gives an effective $\cos\theta \approx 0.5$. The area is then only

$$ A \approx \frac{239}{1361 \times 0.30 \times 0.5} \approx \frac{239}{204} \approx 1.2\ \text{m}^2 . $$

The array is tiny — barely a square meter. The solar collector was never the problem. The problem is storing two weeks of darkness, and no battery technology on the horizon makes $328\ \text{kg}$ for $100\ \text{W}$ acceptable. Solar-plus-battery fails not on generation but on storage mass. We need a source that makes power in the dark.

🧩 Productive Struggle (revisited). This is the problem posed in §25.5. Notice that the failure is specific: it is not that solar power is weak on the Moon (midday flux is higher than on Earth's surface), but that the night is too long to store through. Any world with a long night or a dust-dimmed sky — the Moon, the shadowed poles, Mars in a global dust storm — breaks the store-through-the-dark model. That is precisely the regime the RTG owns.

Phase 3: The radioisotope generator fits the problem exactly

An RTG (§25.4) does not care whether the Sun is up. It makes electricity from radioactive decay, day and night, for decades. Size it: our load is $100\ \text{W}$, and a single MMRTG delivers about $110\ \text{W}$ at beginning of life for a mass of ~$45\ \text{kg}$.

• Power: $110\ \text{W}$ BOL covers the $100\ \text{W}$ load with margin; even after 10 years of decay ($110 \times 0.5^{10/87.7} \approx 101\ \text{W}$ from thermal decay alone) it still meets the load — exactly the end-of-life sizing discipline of §25.2.
• Mass: ~$45\ \text{kg}$, against the solar design's $328\ \text{kg}$ battery. The RTG is about seven times lighter and needs no square meter of deployable array.
• A thermal bonus. The MMRTG rejects roughly $2{,}000\ \text{W}$ of heat to make its 110 W of electricity. On a world that plunges to $-170\,^\circ\text{C}$, that waste heat is not waste at all — it can be plumbed to keep the electronics and battery warm through the night (Chapter 24), solving the survival-heater problem for free. The RTG's greatest inefficiency becomes, in the cold dark, a feature.

The decision is not close. For a low-power station facing the lunar night, the RTG is not merely better — it is the only choice with a believable mass.

📜 From History: This is not hypothetical. Every Apollo surface science station (the ALSEP packages left by Apollo 12 through 17) was powered by a SNAP-27 RTG of about $70\ \text{W}$, precisely because the experiments had to run through the lunar night that would have killed any solar-and-battery design of the era. The astronauts carried the plutonium fuel capsule down separately and inserted it into the generator on the surface. Some of those RTGs powered their instruments for years, until the experiments were switched off from Earth. The lunar night forced nuclear power onto the Moon in 1969, for the same reason it does today.

Phase 4: Scale it up — a crewed base needs fission

Now change one number. A crewed lunar base does not need $100\ \text{W}$; it needs $10\ \text{kW}$ — life support, heating, science, and eventually oxygen production from regolith. Re-run both failing and working options at the new scale.

Battery through the night (for reference): $10{,}000\ \text{W} \times 354\ \text{h} = 3{,}540{,}000\ \text{W·h} = 3.54\ \text{MW·h}$. Even ideally at $150\ \text{W·h/kg}$ that is $23{,}600\ \text{kg}$ — 24 tonnes of battery, more than a fully fueled lunar lander. Storage is out of the question.

RTGs: $10\ \text{kW}$ at ~110 W each would need ~90 MMRTGs and, at ~$4.8\ \text{kg}$ of plutonium oxide each, on the order of hundreds of kilograms of Pu-238 — decades of the entire national production (§25.4). Also out of the question.

Fission (§25.5): a Kilopower-class reactor scaled to $10\ \text{kW}$ masses on the order of $1{,}500\ \text{kg}$ (Tier 2), runs continuously through day and night for a decade or more on a few kilograms of uranium-235, and its Stirling convertors turn heat to electricity at ~30% — five times an RTG's efficiency. At the kilowatt scale, only fission closes the mass budget. This is exactly why NASA's lunar and Mars surface-power plans center on fission reactors, not panels: the two-week night, times ten kilowatts, is a fission problem.

Phase 5: The decision, and the sanity check

Assemble the trade as a table — the kind that ends a real power-subsystem review:

Sanity check by specific power. The lunar night makes storage cost about $150\ \text{W·h/kg} \to$ a brutal $\sim0.4\ \text{W/kg}$ system specific power for solar-plus-battery (328 kg to sustain 100 W). The MMRTG delivers $110/45 \approx 2.4\ \text{W/kg}$, six times better — and fission, at scale, climbs higher still. The ranking by watts-per-kilogram (§25.5) predicts every verdict in the table. The environment chose the power source; we only did the arithmetic.

Discussion Questions

1. The array for the 100 W station came out around $1.2\ \text{m}^2$, yet the design still failed. In one sentence, locate exactly where the solar design breaks and why generation was never the issue.
2. The MMRTG's ~$2{,}000\ \text{W}$ of waste heat was called a "feature" here but a loss in §25.4. Explain how the same quantity can be both, depending on the environment.
3. A colleague proposes solving the lunar-night problem with a much larger array plus batteries, arguing "panels are cheap." Rebut this using the battery-mass number, not the array cost.
4. Why does the case for fission get stronger as the required power rises, while the case for RTGs gets weaker? Tie your answer to Pu-238 supply and to Stirling efficiency.

Your Turn: Extensions

• Option A (design). Re-site the station at a lunar pole, on a ridge in near-permanent sunlight (a "peak of eternal light"). How does the night duration — and therefore the battery — change, and does solar become viable again? What new risk have you taken on?
• Option B (computation). Write night_battery_kg(load_w, night_h, dod, eff, wh_per_kg) and tabulate the battery mass for loads of 50, 100, 500, and 1000 W through a 354-hour night. (Do not run it — hand-trace and add # Expected output:.) At what load does the battery exceed 1 tonne?
• Option C (your mission). If your MDR mission (Track A/B/C/D) has any phase in prolonged darkness or deep space, redo its source selection with this case's logic and record the decision — solar, RTG, or fission — with the mass number that justifies it.

Key Takeaways

1. A long night is a storage problem, and storage is mass. Powering 100 W through the 354-hour lunar night needs ~35 kW·h and ~328 kg of battery — the array (~1.2 m²) was never the issue.
2. RTGs own the low-power, long-dark regime. ~45 kg of MMRTG beats ~328 kg of battery sevenfold and throws in ~2 kW of survival heat for the cold — which is why Apollo's ALSEP ran on SNAP-27 RTGs.
3. At the kilowatt scale, only fission closes the budget. A 10 kW base would need ~24 t of battery or an impossible amount of plutonium; a ~1.5 t Kilopower reactor runs straight through the night.
4. The environment selects the power source; you do the arithmetic. Watts-per-kilogram (§25.5) ranks the options, and the two-week night decides the answer.